Based on snapshots of the servers’ data taken at 3 minute intervals (or less, if I messed up)
Contact me

Server Analysis

First data point: 01 Sep, 12:31:41 PM
Last data point: 04 Sep, 12:27:37 PM

Combined

Most active hours ~ 10AM - 1AM
Least active hours ~ 4AM - 8AM

Dips in the plot are caused by missing data points - a server may not respond when queried for data.
This causes the player count at that point in time to not include an entire server’s worth of players.

Controls
Hover to see the raw data (including full timestamp)
Box select to zoom

Separate

Each server has a slightly different set of operating hours
There are no sharp drops like in the overall line plot because I’ve queried the database for the individual server data points, rather than using query language to simplify the data beforehand

Controls
Click on the red/blue labels on the right to toggle each line
Hover to see the raw data (including full timestamp)
Box select to zoom

Stats

First data point: 30 Aug, 11:25:51 PM
Last data point: 04 Sep, 12:27:37 PM

Maps Played

Of course badwater and upward are the most popular.

Possible improvement to this analysis would be to instead use a value like \[snapshotCount * playerCount\] to more accurately gauge the popularity of each map by considering how many people are actually playing it.
However this seems redundant as the player count is almost always maxed out, so it would be more like \[snapshotCount * 24\] which is kinda useless since there’s no variation

Individual Player Analysis

We can count the playtime of each player.
This assumes every unique name is a separate person. Obviously, this is flawed but whatever.
Values are formatted as hours:minutes:seconds